Warfarin is a drug with historically high adverse drug event (ADE) rates due to its narrow therapeutic/toxicity ratio and variability in inter-individual dose response. Proof of concept testing is proposed to evaluate performance of a novel warfarin dosing approach, which involves modeling of new factors contributing to variability in drug response. This approach will be compared to best current standard of care (i.e.,anti- coagulation clinic algorithms) and assessed by the following primary endpoints: a) weighted measures of % time in therapeutic range as a surrogate for avoided ADE (an intermediate health services research outcome) and b) measured time to stabilization at target INR. The model forms the basis of a dosing calculator which projects stable dose as a function of variability contributed by clinical factors including gender, age, body surface area, diabetes status, heart valves, concomitant use of CYP2C9 inducers, and genetic factors including CYP2C9 and VKORC1 genotype. The model effectively adjusts for 60% of variability in individual patient's warfarin dose. A controlled trial design is planned in which patients are randomized to managed care and empiric dosing using best practices algorithms (N=130), or modeled- calculator based dosing (N=130). Patients will undergo genotyping for CYP2C9 and VKORC1 in real time for the intervention arm to allow genetic variability to be modeled and initial daily dose to be projected prior to drug administration. Standard of care arm will project initial daily dose by applying best practices guidelines. Genotyping will be done for these patients at the end of the study. Subsequent dosing in both arms will be managed in compliance with established dosing algorithms. Patients will be monitored for 2 months post initiation for achievement of primary and additional secondary endpoints. It is proposed that the study will demonstrate that patients can be initiated on warfarin more safely by adjusting for factors contributing to dose response variation thereby promoting more rapid stabilization at target INR. (Lay abstract): A new method to determine the amount of warfarin a patient can safely take is being compared to how the drug is currently prescribed. The new method applies genetic, clinical and environmental data to calculate ideal patient dosing. Calculated dosing is expected to predict safer and more effective treatment.